The production of biotech drugs, pharmaceuticals, neutraceuticals, bio-diesel fuel, as well as many foods and beverages utilizes live cell cultures to implement a biochemical growth process. Optimization of this process during manufacturing requires the ability to control the environment in the bioreactor by detecting a multitude of process variables and controlling their values to be within a specified range of tolerances. Real-time monitoring of these variables and calculations based on these values are performed in order to determine the efficacy of the bioprocess underway.
Recently, systems for controlling process variables applicable to a bioprocess have become increasingly sophisticated. These systems frequently employ digital systems such as programmable logic chips (PLCs), micro-processor based software control systems, or a hybrid arrangement. Advancements in processors, communication hardware, protocols and archival software systems have transformed the concept of data management during bio-processing from a luxury to a necessity. The advent of sophisticated digital systems has given the bio-process engineer the capability to repeatedly apply the same complex series of actions to any bio-process. This has enabled large molecule pharmaceutical manufacturing to move towards the level of reproducibility that semiconductor processing now enjoys. Additionally, the use of digital systems to implement supervisory control and data acquisition (SCADA) now allows a smoother path to satisfying the requirements of good manufacturing process (GMP) doctrines as well as US Food and Drug Administration (USFDA) requirements. However, as capabilities have expanded so have the costs; yet for fully automated control systems and data histories to become commonplace in the biotech industry these process control systems need to be affordable and accessible to even the smallest biotech manufacturing organizations. One route to containing costs is to minimize functional redundancy in the automated process monitoring and control systems.
Irrespective of the complexity of the automation system, each of these control and monitoring platforms for a bioreactor share some degree of commonality. The common elements include a human-machine interface (HMI), a controller, internal and network communications interfaces, instruments by which to measure data from sensors within the bioreactor or adjoining process equipment, and actuators by which to physically interface to process equipment such as agitators, valves, pumps, mass flow controllers (MFCs) and/or rotameters.
A prior art system is shown schematically in FIG. 1:                1.1 is the HMI,        1.2 is a unit that contains the controller and network communications interfaces,        1.3 is a utility tower unit that contains sensor transmitters, relays, and analog outputs for actuator control, as well as electronics that aggregate and condition communications from both transmitters and actuators,        1.4 is a materials handling unit for gases and/or liquids that contains pumps and MFCs,        1.5 is a first bioreactor and,        1.6 is a second bioreactor also coupled to utility tower 1.3        
Optionally, additionally utility towers, pumps and bioreactors (3 and 4) can be operably connected to the same controller and HMI as shown. Note that this architecture can be implemented using either an aggregated design where all of the components of bioreactor units 1 through 4 are packaged in a single enclosure that can control multiple bioreactors (e.g., FIG. 2a , shows an Applikon i-Control unit for control of two bioreactors, where 2.1 is the HMI, 2.2 is a set of integrated pumps, and 2.3 shows two banks of rotameters and FIG. 2b shows several i-Control units in a network).
Alternatively, a modular design with multiple enclosures is possible e.g., as shown in FIGS. 3a and 3b , which illustrates a Finesse TruViu RDPD automation system where 3.1 is the HMI, 3.2 is the utility tower,3.3 is the pump tower, and 3.4 is the MFC unit). The optimal design employed by the end user is influenced by cost, space, maintenance, and ease-of-use requirements for each particular bio-process application. In general, a modular approach offers the greatest flexibility to the end user.
A personal computer (PC) is often used as a terminal or interface through which to access the automation controller and software. The HMI can be a monitor and keyboard that are directly attached to the PC or a separate touch screen display connected using a wireless device. If the PC is used as a terminal, the software values and instructions can reside there, and any executable code is downloaded to the controller where it runs independently of the PC. In cost sensitive applications such as research facilities, where process down-time is less of an issue, the controller can be directly implemented in the PC, whereas in applications requiring high up-time, the controller is often implemented as a separate device to enhance the reliability of the system. The separate controller either has available, or alternatively is packaged with, communications ports for communication with an external network in order to send and receive user commands, instructions, and/or new executable code.
FIG. 4 shows a schematic of a typical utility tower implementation. The utility tower is an enclosure housing the transmitters for the sensors/probes used in a bioprocess. Possible sensors include those which measure:                1. pH,        2. Dissolved oxygen,        3. Pressure,        4. Temperature,        5. Foam level,        6. Liquid level,        7. Weight,        8. Agitator motor speed and/or rocking period and angle,        9. Pump motor speed or number of revolutions, and        10. Gas flow rate        
In general all of these sensors will not utilize the same communication protocol. Some sensors output their signal as a 4 mA to 20 mA analog current, others use HART , while still others use ModBus, (trade mark of Gould Inc Profi Bus, FieldBus, DeviceNet (trade mark of Device Net Vendor Assoc.), Ethernet, wired serial protocols such as RS-232or RS-485 or wireless such as Bluetooth (trade mark of Bluetooth SIG. Inc.) or 802.15 or WiFi 802.11 g. Some sensors use proprietary communication protocols developed by their manufacturer. In order to efficiently send signals to the controller, all or at least most of these sensor signals must be transformed into a common protocol and then aggregated in the utility tower. The aggregated communication line often employs serial communications using a bus. There are many digital bus communication protocols including, but not limited to, ModBus, ProfiBus, DeviceNet, and FieldBus.
FIG. 4 shows a typical prior art utility tower in which 4.1 is a pH transmitter, 4.2 is a dissolved oxygen transmitter, and 4.3 is the communication between these transmitters and their associated sensors, respectively, and 4.4 is the communication between these transmitters and the signal translator. These transmitters often use a HART protocol and therefore need to be sent through a device 4.5 that will translate the HART signals into a suitable Bus protocol, such as ModBus. The signals are then sent to a main signal aggregator 4.6. Many of the other analog or digital signals come in on lines 4.7 from the bioreactor. The pump and MFC towers are conditioned (translated) and aggregated in a separate component 4.8. These signals are then sent on their own line 4.9 to the main signal aggregator 4.6. The totality of the aggregated signals 4.10 is sent to the control tower, to be received by the serial input device. Note that several aggregators and translators can be sequenced, in order to expand the capacity and capability of the overall utility tower system.
Current practice calls for the sensor to connect to the bio-process SCADA system via a transmitter. A typical dissolved oxygen or pH transmitter is shown schematically in FIG. 5 where 5.1 is the transmitter enclosure, 5.2 is the display, 5.3 is the data entry keypad, 5.4 is the cable to the sensor/probe, 5.5 is the sensor/probe, and 5.6 is a data line output from the transmitter that carries the process variable information (and in some cases, additional diagnostic information and/or secondary/tertiary process variables). This data line can be a 4-20 mA analog or HART signal that is physically carried on 2 wires and carries information about the oxygen concentration or pH measured by the probe/transmitter pair. The output signal can also be encoded using a variety of digital protocols (e.g., RS-232, RS-485 etc.) with the value of the process variable contained therein. A transmitter is typically mounted inside a utility (or transmitter) tower, where its output signal is usually aggregated with that of other transmitters and/or actuators, and is sent to the control unit by the signal pathway described previously and illustrated in FIG. 4.
In this scenario, the power (e.g., 24 V DC) to the transmitter is provided by the utility tower, the power to the sensor is provided by the transmitter, and the signal from the sensor is received and conditioned by the transmitter. For transmitters having digital communications capability, i.e., more than just a 4 to 20 mA output signal, the transmitter is subservient to the automation system; namely, the user inputs a command through the automation system's HMI and the sensor transmitter reacts accordingly.
For instance, a typical polarographic dissolved oxygen transmitter or electrochemical pH sensor transmitter will allow the probe to be calibrated and then provide the calibrated probe signal as an output. In this scenario, the transmitter performs these tasks in response to commands sent by the bioprocess automation system. Additionally, as shown in FIG. 4 or FIG. 5, a transmitter with a microprocessor can transmit both the conditioned signal as well as raw signals from the probe. Specifically, digital transmitters are capable of transmitting to the bio-process automation system, both the signals with stored calibrations applied to them, as well as the raw signals (voltage or current or impedance) coming to the transmitters directly from the probes. The raw signals from the probe are important as they can be used in a diagnostic or trouble-shooting capacity. For instance, with a pH probe, the raw output signal is the voltage (mV) developed in accord with the Nernst equation. This raw voltage, when monitored over time, can give insight into the probe's performance (e.g., output range, drift, etc.); additionally the impedance of the probe is indicative of its health (e.g., if the impedance rapidly drops to zero, a probe failure has occurred, and probe output should no longer be trusted as an accurate measurement of pH).
Similar output and diagnostic signals are available from multiple sensors. It should be noted that although these signals are sometimes accessible on the transmitter's display, they are often difficult or impossible to access in a typical bio-process automation system. Similarly, in many inexpensive bio-process automation systems, only the primary process variable from a sensor is measured and converted into a digitized form by proprietary electronics, so that the diagnostic information from raw signal values and/or secondary sensor signals are lost.